The example is first of all for others, and beyond the self. Sometimes, perhaps always, whoever gives the example is not equal to the example he gives, even if he does everything to follow it in advance, "to learn how to live," as we were saying, imperfect example of the example he gives--which he gives by giving then what he has not and even what he is not. For this reason, the example thus disjoined separates enough from itself or from whoever gives it so as to be no longer or not yet example for itself. We do not have to solicit the agreement of Marx--who died to this even before being dead--in order to inherit it: to inherit this or that, this rather than that which comes to us nevertheless by him, through him if not from him. And we do not have to suppose that Marx was in agreement with himself. ("What is certain is that I am not a Marxist," he is supposed to have confided to Engels. Must we still cite Marx as an authority to say likewise?) For Blanchot does not hesitate to suggest that Marx had difficulty living with this disjunction of the injunctions within him and with the fact that they were untranslatable into each other. How is one to receive, how is one to understand a speech, how is one to inherit it when it does not let itself be translated from itself to itself? This may appear impossible. And, we have to acknowledge, it is probably impossible. But since this sums up perhaps the strange subject of this lecture devoted to the specters of Marx, as well as to the avowed distortion of its axiom, permit me then to turn the objection around. Guaranteed translatability, given homogeneity, systematic coherence in their absolute forms, this is surely (certainly, a priori and not probably) what renders the injunction, the inheritance, and the future--in a word, the other--impossible. There must be disjunction, interruption, the heterogeneous if at least there must be, if there must be a chance given to any "there must be" whatsoever, be it beyond duty.

- Derrida, Specters of Marx